[Note for bibliographic reference: Melberg, Hans O. (1997), About reviewing
econometric textbooks, http://www.oocities.org/hmelberg/papers/970425.htm]
About reviewing econometric textbooks
by Hans O. Melberg
This is a short introduction to a five reviews of books on econometrics and statistics.
The purpose of this introduction is to clarify the criteria for the reviews. In other
words: What are the ingredients of a good book on econometrics?
The first criterion, I will argue, is clarity of explanation. In short, to what extent
does to author manage to convey the intuitive idea behind the concepts he is using? This
goal might conflict with another aim - the desire to present rigorous proofs. However,
some books are very good at both, for example by relegating the most complex proofs to
appendicies and footnotes.
A third criterion, is whether the book clearly admits the limitations of statistics.
This is not only an issue of honesty. If the author leaves out important critical
discussions - for example about the arbitrariness of the 5% significance level - the
student may end up believing that statistics is a mechanical process with no room for
critical reflection. This, in turn, may lead to practical problems - such as rejecting a
hypothesis which is very convincing, but not statistically significant. (See my article
Bees, Marbles and Boeing for several examples of this)
In sum, I shall judge the books by these criteria: Clarity of explanation, rigour,
completeness, and honesty about the limitations of econometrics. In addition to these
criteria, of course, I apply the normal stylistic criteria - such as whether the book is
interesting and appealing.
[Note for bibliographic reference: Melberg, Hans O. (1997), About reviewing
econometric textbooks, http://www.oocities.org/hmelberg/papers/970425.htm]